Copy number variations (CNVs) confer significant effects on genetic innovation and phenotypic variation. Previous CNV studies in swine seldom focused on in-depth characterization of global CNVs. Using whole-genome assembly comparison (WGAC) and whole-genome shotgun sequence detection (WSSD) approaches by next generation sequencing (NGS), we probed formation signatures of both segmental duplications (SDs) and individualized CNVs in an integrated fashion, building the finest resolution CNV and SD maps of pigs so far. We obtained copy number estimates of all protein-coding genes with copy number variation carried by individuals, and further confirmed two genes with high copy numbers in Meishan pigs through an enlarged population. We determined genome-wide CNV hotspots, which were significantly enriched in SD regions, suggesting evolution of CNV hotspots may be affected by ancestral SDs. Through systematically enrichment analyses based on simulations and bioinformatics analyses, we revealed CNV-related genes undergo a different selective constraint from those CNV-unrelated regions, and CNVs may be associated with or affect pig health and production performance under recent selection. Our studies lay out one way for characterization of CNVs in the pig genome, provide insight into the pig genome variation and prompt CNV mechanisms studies when using pigs as biomedical models for human diseases.

Background: In broiler production, breast muscle weight and intramuscular fat (IMF) content are important economic traits. Understanding the genetic mechanisms that underlie these traits is essential to implement effective genetic improvement programs. To date, genome-wide association studies (GWAS) and gene expression analyses have been performed to identify candidate genes for these traits. However, GWAS mainly detect associations at the DNA level, while differential expression analyses usually have low power because they are typically based on small sample sizes. To detect candidate genes for breast muscle weight and IMF contents (intramuscular fat percentage and relative content of triglycerides, cholesterol, and phospholipids), we performed association analyses based on breast muscle transcriptomic data on approximately 400 Tiannong partridge chickens at slaughter age. Results: First, by performing an extensive simulation study, we evaluated the statistical properties of association analyses of gene expression levels and traits based on the linear mixed model (LMM) and three regularized linear regression models, i.e., least absolute shrinkage and selection operator (LASSO), ridge regression (RR), and elastic net (EN). The results show that LMM, LASSO and EN with tuning parameters that are determined based on the one standard error rule exhibited the lowest type I error rates. Using results from all three models, we detected 43 candidate genes with expression levels that were associated with breast muscle weight. In addition, candidate genes were detected for intramuscular fat percentage (1), triglyceride content (2), cholesterol content (1), and phospholipid content (1). Many of the identified genes have been demonstrated to play roles in the development and metabolism of skeletal muscle or adipocyte. Moreover, weighted gene co-expression network analyses revealed that many candidate genes were harbored by gene co-expression modules, which were also significantly correlated with the traits of interest. The results of Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses indicated that these modules are involved in muscle development and contraction, and in lipid metabolism. Conclusions: Our study provides valuable insight into the transcriptomic bases of breast muscle weight and IMF contents in Chinese indigenous yellow broilers. Our findings could be useful for the genetic improvement of these traits in broiler chickens.

As a major component of genomic variation, copy number variations (CNVs) are considered as promising markers for some phenotypic and economically important traits in domestic animals. Using a custom-designed 1M array CGH (aCGH), we performed CNV discovery in 12 pig samples from one Asian wild boar population, six Chinese indigenous breeds, and two European commercial breeds. In total, we identified 758 CNV regions (CNVRs), covering 47.43 Mb of the pig genome sequence. Of the total porcine genes, 1295 genes were completely or partially overlapped with the identified CNVRs, which enriched in the terms related to sensory perception of the environment, neurodevelopmental processes, response to external stimuli, and immunity. Further probing the potential functions of these genes, we also found a suite of genes related important traits, which make them a promising resource for exploring the genetic basis of phenotype differences among diverse pig breeds. Compared with previous relevant studies, the current study highlights that different platforms can complement each other, and the combined implementation of different platforms is beneficial to achieve the most comprehensive CNV calls. CNVs detected in diverse populations herein are essentially complementary to the CNV map in the pig genome, which would be helpful for understanding the pig genome variants and investigating the associations between various phenotypes and CNVs.

A lack of the complete pig proteome has left a gap in our knowledge of the pig genome and has restricted the feasibility of using pigs as a biomedical model. In this study, we developed a tissue-based proteome map using 34 major normal pig tissues. A total of 5841 unknown protein isoforms were identified and systematically characterized, including 2225 novel protein isoforms, 669 protein isoforms from 460 genes symbolized beginning with LOC, and 2947 protein isoforms without clear NCBI annotation in the current pig reference genome. These newly identified protein isoforms were functionally annotated through profiling the pig transcriptome with high-throughput RNA sequencing of the same pig tissues, further improving the genome annotation of the corresponding protein-coding genes. Combining the well-annotated genes that have parallel expression pattern and subcellular witness, we predicted the tissue-related subcellularlocations and potential functions for these unknown proteins. Finally, we mined 3081 orthologous genes for 52.7% of unknown protein isoforms across multiple species, referring to 68 KEGG pathways as well as 23 disease signaling pathways. These findings provide valuable insights and a rich resource for enhancing studies of pig genomics and biology, as well as biomedical model application to human medicine.

Genomic selection (GS) is an efficient method to improve genetically economic traits. Feature selection is an important method for GS based on whole-genome sequencing (WGS) data. We investigated the prediction performance of GS of milk production traits using imputed WGS data on 7957 Chinese Holsteins. We used two regularized regression models, least absolute shrinkage and selection operator (LASSO) and elastic net (EN) for feature selection. For comparison, we performed genome-wide association studies based on a linear mixed model (LMM), and the N single nucleotide polymorphisms (SNPs) with the lowest p-values were selected (LMM_LASSO and LMM_EN), where N was the number of non-zero effect SNPs selected by LASSO or EN. GS was conducted using a genomic best linear unbiased prediction (GBLUP) model and several sets of SNPs: (1) selected WGS SNPs; (2) 50K SNP chip data; (3) WGS data; and (4) a combined set of selected WGS SNPs and 50K SNP chip data. The results showed that the prediction accuracies of GS with features selected using LASSO or EN were comparable to those using features selected with LMM_LASSO or LMM_EN. For milk and protein yields, GS using a combination of SNPs selected with LASSO and 50K SNP chip data achieved the best prediction performance, and GS using SNPs selected with LMM_LASSO combined with 50K SNP chip data performed best for fat yield. The proposed method, feature selection using regularization regression models, provides a valuable novel strategy for WGS-based GS.